Skip to content

Signal processing pipeline to perform preprocessing (including motion and muscle artifact removal), dimensionality reduction, and source localization from electroencephalography data, and visualize changes in gait-related spectral power

License

Notifications You must be signed in to change notification settings

jacobsen-noelle/ExoAdapt-DualEEG-Processing

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ExoAdapt-DualEEG-Processing

This pipeline performs preprocessing on dual-electrode electroencphalography data and source analysis to investigate gait-related changes in electrocortical activity during adaptation to a powered ankle exoskeleton. All local data processing is written for MATLAB and relies heavily on the EEGLAB toolbox (see Toolboxes.txt for additional external tools or EEGLAB plug-ins used). Some of the steps (independent component analysis, custom head modeling, source localization) are designed to be used with supercomputing resources.

These are the steps for dual-EEG processing:

  1. Dual-EEG Preprocessing This step uses dual-electrode EEG and neck electromyography to isolate and remove motion and muscle artifacts from EEG [1,2]. It also rejects noisy channels and time windows.
  2. Add AMICA results After EEG preprocessing, Adaptive Mixture Indpendent Component Analysis is run on supercomputing resources extract independent components from "clean" scalp EEG data (see AMICA folder). This steps pulls results from the supercompter and adds the ICA matrix to the EEG dataset.
  3. Add DIPFIT results Source localization using DIPFIT and custom head models is performed using supercomputing resources. The custom-head-model folder provides a pipeline for creating 5-layer volume conducting head models with the finite element method, following the fieldtrip-SIMBIO pipleine. The source-localization folder provides code for running DIPFIT with the custom head models and implements parallel processing to improve computation speed. This step adds source model results from DIPFIT located on the supercomputer back into the local EEG dataset
  4. IC rejection Automatic rejection of non-brain independent components based on a spatial and temporal features
  5. Epoching and Timewarping Epochs and timewarps data based on gait events
  6. Group Analysis Precomputes measures (e.g. power spectral density, event-related spectral pertubations) and performs component clustering
  7. Plot results Perform statistical analyses and plot results

[1] N. Richer, R. J. Downey, W. D. Hairston, D. P. Ferris, and A. D. Nordin, “Motion and Muscle Artifact Removal Validation Using an Electrical Head Phantom, Robotic Motion Platform, and Dual Layer Mobile EEG,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 28, no. 8, pp. 1825–1835, Aug. 2020, doi: 10.1109/TNSRE.2020.3000971. [2] Ryan J Downey, Daniel P Ferris, "iCanClean Removes Motion, Muscle, Eye, and Line-Noise Artifacts from Phantom EEG," Sensors. Oct. 2023

About

Signal processing pipeline to perform preprocessing (including motion and muscle artifact removal), dimensionality reduction, and source localization from electroencephalography data, and visualize changes in gait-related spectral power

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published